Investigating Visual System Neurons in Drosophila melanogaster through Machine Learning

Overview

Janelia Research Campus The visual system is the component of the nervous system that processes visual information. Scientists understand that all neuron types in the human visual system work together to analyze visual information, which leads to an approach or escape from a visual stimulus. However, they have a limited understanding as to how each individual neuron type affects behavior. The human visual system is complex, as it contains over 140 million neurons that aid in visual processing. The fruit fly, Drosophila melanogaster , is therefore a model organism to study the visual system, since its compact genome shares 60% of genes with that of humans and can be easily manipulated. Genetic driver lines were used to silence lobula plate visual system neuron groups H2 and HS in flies. These flies were placed in a fly arena and were subject to 18 moving light stimuli. Each fly experiment was recorded, and a Mixture Density Network algorithm was used to track the flies’ head movements in response to the stimuli. The output was then uploaded onto MATLAB to calculate for the angle between the flies’ head and body. The movements of the wild type flies and flies that have silenced H2 and HS neurons was compared. Flies that had silenced H2 and HS neurons had slower reaction times to the stimuli, and more staggered head movements. The data also show that the H2 neuron affects clockwise head motion more than counterclockwise movements. These results are novel and are being confirmed with fly body movement data.

Competition history

• JSHS 2020 Category not listed

Resources

• National JSHS abstract book ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.